| Summary: | As the processing power and the science of robotics advances, robots are becoming more and more available to the masses and convenient to use. One instance of the applications of robots are the moving robots. These robots essentially move and explore the world they are located in, mainly: indoor robots. In our laboratory we are currently working on an indoor robot which serves the purpose of teleperesence. One of the challenges in building such robot is the navigation, mainly planning routes, and avoiding the obstacles that are in front of it. In order to navigate, the robot needs to have a map of the environment where it is located. Plus, it needs to locate itself in that map. This paper discusses an improved obstacle avoidance algorithm to our laboratory's experimental telepresence robot. The improved algorithm is based on Simultaneous Localization and Mapping (SLAM) This new obstacle avoidance algorithm uses a Kinect 2 to collect depth frames and RGB images then synthesize them into Point Cloud files (PCL). These Point clouds are processed and gathered into a 3D map and used as input to the Simultaneous Localization and Mapping (SLAM) algorithm to help the robot navigate as smooth as possible in its indoor environment. The robot navigates by creating a map of its surroundings and constantly localizing itself in that map. © 2016 IEEE.
|
|---|